Stock prices don't move in isolation. When SAP announces earnings, Oracle and Microsoft move too. When the entire tech sector rallies, your stock might rise for reasons that have nothing to do with the company itself.

FIN lets you define competitor tickers for any of your stocks. When competitors are configured, the prediction model sees the full competitive landscape — not just your stock's data, but how it stacks up against its peers.

Why Add Competitors?

Without Competitors	With Competitors
The model only sees your stock's price action and news	The model sees relative performance — is my stock beating or lagging its peers?
Can't distinguish sector trends from company-specific moves	Can tell the difference: "all tech stocks are down" vs "only my stock is down"
No context on valuation (is a high P/E normal for this sector?)	Comparative valuation shows whether your stock is cheap or expensive relative to peers
News analysis limited to your ticker's articles	Competitor news sentiment and surprises are also factored in

Concrete Examples

• Sector-wide rally: If your stock is up 3% but all competitors are up 3-5%, the model learns this is sector strength — less likely to reverse than a company-specific surge
• Relative weakness: If your stock rises on positive news but rises less than competitors, the model flags this as a potential weakness signal
• Divergence: If your stock drops while competitors rise, the model looks for company-specific explanations rather than assuming a market downturn

How It Works

Adding Competitors

Go to the dashboard's Competitors page. For any ticker you're tracking, you can add competitor tickers:

Ticker: SAP.DE
Competitors: ORCL, CRM, NOW

When you add a competitor, FIN automatically generates RSS news feeds for that competitor (using the same feed templates as your primary ticker). Competitors that aren't also standalone tickers in your account receive RSS-feed-only scheduling — they collect news but don't run their own predictions.

What Gets Collected

Once competitors are configured, FIN collects data from three sources:

1. Price Performance (from Yahoo Finance)

• 5-day and 20-day returns for your stock and each competitor
• A relative strength assessment: is your stock outperforming, underperforming, or in line with peers?

2. Financial Metrics (from Yahoo Finance)

• P/E ratio and forward P/E
• Revenue growth and profit margin
• EV/EBITDA

These are presented in a comparative valuation table with peer medians, so the model can spot valuation premiums or discounts.

3. News Sentiment & Surprises (from ChromaDB)

• Recent news headlines for each competitor with their sentiment scores
• Aggregated average sentiment per competitor
• Surprise signals (earnings beats/misses, guidance changes, economic surprises)

This means competitor news isn't just collected — it's analyzed through the same sentiment, surprise, and psychology pipeline as your primary ticker's news, and the results are included in the prediction prompt.

How It Affects Predictions

When the prediction runs, a self-contained === Competitor Analysis === section is injected into the LLM prompt containing:

RELATIVE PRICE PERFORMANCE:
  SAP.DE: 5d=-1.2% 20d=+3.4%
  ORCL:   5d=+0.8% 20d=+5.1%
  CRM:    5d=-0.5% 20d=+2.1%
  → SAP.DE UNDERPERFORMING peers by -1.7pp

COMPARATIVE VALUATION:
  Ticker       P/E  Fwd P/E  Rev Growth   Margin  EV/EBITDA
  SAP.DE     28.5     26.3       11.0%    18.2%      22.1
  ORCL       32.1     29.8        8.5%    22.5%      26.3
  CRM        24.8     23.1       12.2%    16.8%      20.5
  Peer median P/E: 28.5
  → SAP.DE trades in line with peers

COMPETITOR NEWS SENTIMENT:
  ORCL: avg_ticker_sentiment=62.0, articles=3
    "Oracle Cloud Revenue Surges" (ticker_sentiment: 65)
    "Oracle AI Partnership Announced" (ticker_sentiment: 70)
  CRM: avg_ticker_sentiment=45.0, articles=1
    "Salesforce Faces Analyst Downgrade" (ticker_sentiment: 38)

COMPETITOR SURPRISE SIGNALS:
  ORCL: 1 surprise(s) (1 positive, 0 negative, avg_magnitude=18.0)

The model uses this data to:

• Determine whether price moves are sector-wide or company-specific
• Assess relative strength as a potential signal
• Compare valuation to identify premium/discount situations
• Factor in competitor news sentiment when forming its thesis

Managing Competitors

Dashboard UI

The Competitors page lets you:

• View all competitors for any of your tickers
• Add new competitors (enter ticker symbol and optional description)
• Toggle competitors on/off without deleting them
• Remove competitors (also cleans up auto-generated RSS feeds)

Configuration Options

Competitors can be disabled per ticker without removing them. Go to the Prompt Config page and set competitor_context: {enabled: false} for any ticker where you want to suppress competitor data from the prediction prompt.

Things to Know

• Competitor RSS feeds need time to accumulate data — on first setup, there won't be any news or surprise data until the feeds have been polled at least once
• Competitor-only tickers don't generate predictions — they're purely data sources for your primary ticker's analysis
• Competitor data is optional — if you don't configure any competitors, the prediction flow is unchanged; the feature simply doesn't activate
• Competitor news is analyzed using LLM calls — adding competitors increases LLM usage proportionally to the number of competitor articles processed